Overall, I’m very surprised at how little media coverage there’s been around the OpenAI / Hugging Face attack. It’s clearly one of the most important things to happen this year.
I’ve joined Cursor / SpaceXAI.
AI is bound by compute, and SpaceXAI has the best near and long term compute roadmap of any AI lab.
Cursor + SpaceX as a joint entity is what sold me: Cursor’s product taste, coding data, and stellar GTM are now backed by the largest open reserve of compute on earth.
The longer term roadmap is even stronger: SpaceX will have by far the lowest launch cost for orbital compute. And Terafab provides dedicated capacity with a domestic supply chain.
I’ve long wanted to work alongside the model team. I’m thrilled and honored to do this work at SpaceXAI.
Thanks @romanugarte_@CathieDWood@mntruell
@paulg Of course, you know a tiny handful of my 10,000+ employees aren't citizens. I am pro-American, not the strict polar opposite of your traitorous ideology.
And I won't play 20 Questions with you as you try to bait me into saying something we both know is against California law.
The Robotics team from Wissahickon High School in Ambler, Pennsylvania, built the robot Miss Daisy XXIV that picks up balls and shoots them into a container.
Most common reasons for failing in order:
1) Never starting
2) Quitting too soon
3) Not doing enough volume
4) Not doing it consistently enough
5) Doing too many different things
6) Not shooting big enough
a surprise I think is coming
the quick payback periods and extraordinarily high IRR for AI infrastructure, even at monumental scale, will push up cost of capital sufficient to push many traditional businesses into the abyss, even those without obvious direct AI counter-exposure
🇻🇪 Watch Venezuela.
Bloomberg reports Venezuela is considering leaving OPEC.
But don’t watch OPEC in isolation.
Watch the gold too.
Venezuela is trying to recover roughly 31 tonnes held at the Bank of England.
Ignore reports claiming it has already been transferred to the US. That has not been established.
The interesting part is what happens next.
Venezuela could become a perfect example of the MIC → FIC → TIC transition I’ve been describing.
The MIC does the dirty work.
The operation against Maduro was about changing the political alignment and pulling Venezuela back into the FIC sphere.
Then the FIC takes over.
The gold gets negotiated.
The sanctions get negotiated.
Access to financial markets gets negotiated.
The oil gets negotiated.
And the currency war begins.
If Venezuela leaves OPEC, regains access to its frozen assets and keeps its oil inside dollar settlement, the FIC doesn’t need the old petrodollar system in the same way.
Especially if the TIC gets the final piece:
Stablecoins.
Move Venezuelan oil and commerce onto dollar stablecoin rails and you have a new model.
MIC changes the regime.
FIC controls the money.
TIC controls the rails.
That’s the transition.
The US can shrink from global hegemon to regional power of the Western Hemisphere only.
The less likely outcome is if Venezuela leaves OPEC, gets its gold back and starts settling more oil outside the dollar, then the story changes.
That would tell me Venezuela is moving towards the multipolar financial system rather than being absorbed into the US-led node.
That’s why the OPEC headline is only half the story.
Saudi Arabia is already connected to mBridge, infrastructure that can settle cross-border, multi-currency payments without relying on the traditional correspondent banking chain.
The old architecture is being dismantled.
The new nodes are being built.
America consolidates its node.
China consolidates its node.
Everyone else hedges.
And the FIC/TIC fight over whose currency and rails connect them.
MIC → FIC → TIC.
That tells you what Venezuela was really about.
Holy shit! This might be one of the most underrated releases of the year.
Anthropic just gave Claude the ability to control physical lab equipment like microscopes, robotic arms, liquid handlers, lasers.
The results are wild:
– QuEra had Claude fix quantum computer lasers overnight, unsupervised. A fix that took human experts 5-10 minutes, Claude got down to 6 seconds. Success rate: 58% → 99.3%.
– Genentech had Claude self-optimize lab pipetting by scoring its own results against expert baselines, and converged on the right settings on its own.
– A university lab went from manual 4 am plate-swapping to Claude Code running the whole workflow hands-free.
– Claude even ran qPCR experiments that helped detect human sewage contamination in a California creek.
Automated science labs are closer than most people think.
Anthropic is somewhat back!
The founder of Shopify builds every product around ONE line:
"Don't make better cameras, make better photographers."
Tobi Lütke borrowed the idea from Kathy Sierra, and he says it sits deep in his psyche.
A great tool should not just work well, it should make the person using it better.
He says beautiful products can pull more ambition out of people, and push them to build skills they never would have otherwise.
That is the whole point of what he makes.
The product is a way to help you become the best version of yourself - the best entrepreneur in ecommerce.
— Tobi Lütke (.@tobi), founder of Shopify
“I use AI to write like I use my calculator to do math” fundamentally misunderstands the human desire for truth seeking.
Calculator math is close-ended, with empirically correct answers. Writing is open-ended, with subjectively nuanced answers. The determination for authenticity in writing, unlike math, is not the actual answer, but the quest it takes.
In other words, if you use AI to write, you are not using a calculator. You are becoming the calculator.
AI is not one technology.
It’s an entire ecosystem.
And the deeper you go, the more you realize that ChatGPT or Claude is only one layer of a much bigger stack.
At the foundation, you have:
→ Machine learning
→ Neural networks
→ Transformers
→ Computer vision
→ Large language models
Then comes the infrastructure that makes these models useful:
→ Data & knowledge systems
→ Fine-tuning
→ Prompt engineering
→ RAG
→ Knowledge graphs
→ Model serving & inference
But that’s still not the end.
The next layer is where AI starts doing things:
→ AI agents
→ Tool orchestration
→ Multi-agent systems
→ Autonomous execution
→ Memory systems
→ Self-reflection & error recovery
And then AI moves into real-world applications:
→ AI copilots
→ Personalization
→ Content generation
→ Code generation
→ Voice & video
→ Business automation
But there’s one part that becomes more important as AI gets more autonomous:
Trust.
Security.
Privacy.
Bias & fairness.
Explainability.
Human oversight.
Risk management.
Governance.
Compliance.
Because building a powerful AI system is only half the problem.
The bigger challenge is building one that people can trust, control and deploy safely.
That’s why the future of AI isn’t just about better models.
It’s about the entire ecosystem working together.
Models + Data + Infrastructure + Agents + Applications + Governance.
Most people are focused on the model.
The real opportunities may be hiding in everything around it.
Save this AI ecosystem map. It’s worth revisiting as the industry evolves.
Follow @AamirAnsar94694 for more practical AI, tech & business insights.
#AI #ArtificialIntelligence #GenerativeAI #AIAgents #MachineLearning #LLM #AIEngineering #FutureOfAI